Presentation 2002/9/21
Categorical judgments of facial expressions by RBF network simulation
Takuto YAMAGUCHI, Shinichi ASAKAWA, Hiroshi YAMADA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) It can be postulated a hypothesis that the categorical judgments of facial expressions are based on the probability density functions of emotional categories in the multidimensional space of physical attributes. The present study attempted to test it by comparing a neural network simulation with a psychological experiment data. In the neural network simulations, we created a 3-layered radial basis function (RBF) network, which has Gaussian basis function in hidden units. The network learned 216 schematic facial expressions of 6 emotions which were obtained from Yamada (1993). As a result, the simulations could judge facial expressions correctly (90-100% correct answers). And we obtained high canonical correlations between the network's performance and the participants' judgment distributions. Thus it can be concluded that the hypothesis described above is supported.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) categorical judgments of facial expressions / RBF network / probability density functions
Paper # HCS2002-27
Date of Issue

Conference Information
Committee HCS
Conference Date 2002/9/21(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Human Communication Science (HCS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Categorical judgments of facial expressions by RBF network simulation
Sub Title (in English)
Keyword(1) categorical judgments of facial expressions
Keyword(2) RBF network
Keyword(3) probability density functions
1st Author's Name Takuto YAMAGUCHI
1st Author's Affiliation Graduate School of Literature and Social Sciences, Nihon University()
2nd Author's Name Shinichi ASAKAWA
2nd Author's Affiliation Center for Information Science, Tokyo Women's Christian University
3rd Author's Name Hiroshi YAMADA
3rd Author's Affiliation College of Humanities and Sciences, Nihon University
Date 2002/9/21
Paper # HCS2002-27
Volume (vol) vol.102
Number (no) 342
Page pp.pp.-
#Pages 4
Date of Issue